Magnetic and magnetocaloric properties of Pr2-xNdxFe17 Magnetic and magnetocaloric properties of Pr2-xNdxFe17

The structural, magnetic and magnetocaloric properties of Fe deﬁcient Pr 2-x Nd x Fe 17 (x = 0.5, 0.7) alloys prepared by arc-melting and melt-spinning have been investigated. The room temperature x-ray diffraction patterns show that the samples are nearly single-phase and crystallize in the rhombohedral Th 2 Zn 17 -type crystal structure. The Curie temperatures determined from the thermomagnetic curves are 302 K and 307 K for Pr 1.5 Nd 0.5 Fe 17 and Pr 1.3 Nd 0.7 Fe 17 , respectively. The peak magnetic entropy change and the relative cooling power at ﬁeld change of 50 kOe are 3.01 J/kgK and 345 J/kg for Pr 1.5 Nd 0.5 Fe 17 , and 4.31 J/kgK and 487 J/kg for Pr 1.3 Nd 0.7 Fe 17, respectively . The absence of magnetic and thermal hysteresis with relatively high cooling efﬁciency suggests that the alloys have potential for magnetic refrigeration.


I. INTRODUCTION
Magnetic refrigeration is expected to become an alternative to the current gas-compression cooling system with advantages such as high energy efficiency, low manufacturing and maintenance costs, and use of environment friendly solid state materials. [1][2][3][4][5][6] The magnetic refrigeration (MR) is based on the magnetocaloric effect (MCE), where a magnetic material shows thermal response to an external magnetic field. 7 The MCE is measured either in terms of the magnetic entropy change (∆SM) in an isothermal process or in terms of the temperature change (∆T ad ) in an adiabatic process. Therefore, development of magnetic materials with large ∆SM and ∆T ad over a broad temperature range near room temperature is crucial for the advancement of magnetic refrigeration technology. 8 The value of ∆SM depends on the rate at which magnetization changes with temperature (∂M/∂T). Therefore, large values of ∆SM are obtained near phase transitions. Materials undergoing magnetostructural phase transitions (first-order phase transition FOPT), such as Gd-Si-Ge, 9 La-Fe-Si, 10 Mn-Fe-P-As, 11 and Ni-Mn-based Heusler alloys 12 exhibit very large values of ∆SM near room temperature due to sharp changes in magnetizations. However, materials undergoing FOPT are not considered ideal for magnetic refrigeration because of large magnetic and thermal hysteresis, narrow operating temperature range, irreversibility in entropy and temperature changes over magnetization/demagnetization cycles, and issues with mechanical stability. 13 Most of these issues can be addressed by using materials that show second-order phase transitions (SOPT) as magnetic refrigerants, but they exhibit relatively lower values of ∆SM. Among SOPT materials, elemental gadolinium (Gd) combines moderate value of ∆SM with relatively high cooling efficiency but it is expensive and highly susceptible to oxidation. 3 Other SOPT materials yielding entropy changes comparable to Gd are R 2 Fe 17 -type (R = rare-earth element) alloys 14 which contain much smaller amounts of expensive rare earths. Our focus is on the Curie temperatures TC of these alloys, which can be fine-tuned by adjusting the elemental composition, especially through the choice of rare-earth combinations.
Here we investigate the structural, magnetic and magnetocaloric properties of Fe deficient Pr 2-x NdxFe 17 16,17 Since the Tc of Pr 2 Fe 17 is somewhat lower than room temperature (283 K) and that of Nd 2 Fe 17 is somewhat higher than room temperature (340 K), it is possible to adjust the Tc of Pr 2-x NdxFe 17 near room temperature by adjusting the elemental composition. We have found that the Fe deficient Pr 1.3 Nd 0.7 Fe 17 alloy maintains high magnetization with a Tc of 307 K. In addition to choosing a combination of Pr and Nd as the rare-earths to tune Tc, our experimental investigations suggested that slightly Fe-deficient composition facilitated better phase purity and slightly higher net magnetization in comparison to (Pr,Nd) 2 Fe 17 counterparts, and will be beneficial for the magnetocaloric properties.

II. EXPERIMENTAL METHODS
The Fe deficient Pr 2-x NdxFe 17 (x = 0.5, 0.7) alloys in the form of ribbons were synthesized using arc-melting, followed by rapid quenching in a melt-spinner. First, pieces of Pr, Nd, and Fe with proper weight ratio were cut from commercially available pellets and then melted on a Cu hearth of an arc-melting furnace in a highly pure argon environment. The arc-melted ingots were broken into pieces, induction melted in a quartz tube inside a meltspinner chamber and rapidly quenched onto the surface of a copper wheel rotating at 25 m/s. The phase formation and structural properties of the samples were investigated using powder x-ray diffraction (XRD) using a PANalytical Empyrean Diffractometer with Cu Kα radiation (wavelength of 1.5418 Å). The XRD patterns were analyzed by Rietveld method using TOPAS software (Bruker, AXS). The magnetic properties were investigated using Quantum Design VersaLab magnetometer and Physical Property Measurement System (PPMS). The estimated elemental compositions were confirmed using energy-dispersive x-ray spectroscopy (EDX) in a FEI Nova NanoSEM450. Figure 1 shows the XRD patterns of the rapidly quenched Pr 1.3 Nd 0.7 Fe 17 and Pr 1.5 Nd 0.5 Fe 17 ribbons measured at room temperature. In order to determine the lattice parameters and the presence of secondary phase, the experimental XRD patterns were compared with the simulated patterns by Rietveld refinement method using TOPAS software. The powder x-ray diffraction simulations indicate that the compounds are formed nearly single phase with Th 2 Zn 17 -type rohombohedral crystal structure (R3m space group). Traces of upto 5 wt. % of α-Fe impurity was detected in the phase analysis by Rietveld method. The lattice parameters are: a = 8.579 Å, c = 12.486 Å and a = 8.575 Å, c = 12.499 Å, respectively for x = 0.5 and 0.7, thus indicating only a minor cell expansion along the c-axis in the Nd-rich sample. In the Th 2 Zn 17 -type structure, Pr and Nd share the Wyckoff position 6c (0, 0, 0.333), whereas Fe is distributed among sites 6c (0, 0, 0.097), 9d (0.5, 0, 0.5), 18f (0.333, 0, 0) and 18h (0.5, 0.5, 0.167). The elemetal compositions determined using EDX spectroscopy are Pr 1.3 Nd 0.7 Fe 15 and Pr 1.5 Nd 0.5 Fe 15 and therefore we investigated the Fe-deficiency in the diffraction profile analysis. The analysis suggests that 6c and 18f sites are the Fe-lean sites and the composition based on this analysis matches closely with the nominal composition obtained by EDX spectroscopy. Figure 2 shows the thermomagnetic curves M(T) of the Pr 2-x NdxFe 17 ribbons measured at 1 kOe. For these measurements, the ribbons were first cooled to 100 K at H = 0 and measurements were performed at 1 kOe while the temperature increased from 100 to 380 K at 2 K/min (zero field cool measurement).

III. RESULTS AND DISCUSSION
The M(T) curves are smooth and show a SOPT from a ferromagnetic to a paramagnetic phase. No thermal hysteresis was observed when the measurement was reversed from 380 K to 100 K (not shown). The Curie temperatures obtained from the peak values of dM dT vs. T curves are 302 K for Pr 1.5 Nd 0.5 Fe 17 and 307 K for We have investigated the magnetocaloric properties of Pr 2-x NdxFe 17 (x = 0.5, 0.7) ribbons in terms ∆SM and relative cooling power (RCP). In order to determine ∆SM for the ribbons, we recorded isothermal magnetization curves between 0 and 50 kOe, at various temperatures near their Curie temperatures. The M(H) curves recorded for Pr 1.3 Nd 0.7 Fe 17 and Pr 1.5 Nd 0.5 Fe 17 at various temperature intervals (2 K, 5 K, and 10 K) starting from 2 K near T C are shown in Figs. 3(a) and 3(c), respectively. The entropy changes ∆SM at various temperatures were obtained using Maxwell's thermodynamic relation SM(T, H) = ∫ H 0 dM dT H dH utilizing the M(H) data. Figures 4(a) and 4(b) show the entropy changes ∆SM as a function of temperature. The peak values of magnetic entropy change ∆SM,max at magnetic field change of 50 kOe are 4.31 Jkg -1 K -1 and 3.01 Jkg -1 K -1 for Pr 1.3 Nd 0.7 Fe 17 and Pr 1.5 Nd 0.5 Fe 17 alloys, respectively. These values are comparable to those in other Fe-rich rare-earth intermetallics 14 while the adjusted Curie temperature near 300 K is highly beneficial for room-temperature magnetic refrigeration. Note that R 2 Fe 17 alloys are expected to have similar peak entropies across the lanthanide series, because the relatively weak rare-earth iron intersublattice exchange 19 strongly reduces the rare-earth entropy change near the Curie temperature. By contrast, the Curie temperature increases systematically towards the middle of the lanthanide series (476 K for Gd 2 Fe 17 ).
From a practical viewpoint, the cooling efficiency of a magnetocaloric material is parameterized by the relative cooling power (RCP). The RCP measures the amount of heat transferred from a hot reservoir to cold reservoir in a thermodynamic cycle. The insets of Figs

IV. CONCLUSIONS
In summary, we have prepared Fe deficient Pr 1.3 Nd 0.7 Fe 17 and Pr 1.5 Nd 0.5 Fe 17 alloys using arc-melting and melt-spinning. The x-ray diffraction analysis show that both alloys crystallized in the rhombohedral Th 2 Zn 17 -type structure. The peak magnetic entropy changes of 3.01 J/kgK in Pr 1.5 Nd 0.5 Fe 17 and 4.31 J/kgK in Pr 1.3 Nd 0.7 Fe 15 are comparable to those of other materials with second-order phase transitions and the respective Curie temperatures, 302 K and 307 K, are very close to room temperature. Furthermore, a relatively high cooling power of 487 J/kg has been achieved in Pr 1.3 Nd 0.7 Fe 17 . The peak entropy change, the cooling power, the operation near room temperature, and the relatively low raw-materials price make the alloys potential candidates for room-temperature magnetic refrigerants.